What it is
An awesome-list style repository of 50 real-world use cases for OpenClaw, an open-source personal AI agent. It has 23 China-ecosystem use cases and a set of general international ones, some with China adaptations. Each use case is its own file covering the pain point, what it does, required skills, setup with copyable prompts, and practical tips. It also includes an AGENTS.md / INDEX.md / CONCEPT-MAPPING.md layer so other AI agents can reproduce the use cases.
Who it's for
- Chinese-speaking users who want to start with OpenClaw and need concrete scenarios
- Users who want to connect AI to Feishu, DingTalk, WeCom, Xiaohongshu or WeChat Official Accounts
- People automating office work, content creation, research, monitoring, DevOps or knowledge management
- Users of other AI agents (Claude Code, Codex, Hermes) who want use-case guides they can adapt
Requirements
Requirements
- OpenClaw installed (github.com/openclaw/openclaw)
- Optionally a connected channel such as Telegram, Discord, Feishu, DingTalk or WeCom, depending on the use case
- Skills or plugins listed in each use case; the README says third-party dependencies are not reviewed by the maintainers
- Higher-difficulty (⭐⭐⭐) use cases need some technical background
Setup
Install OpenClaw
Install OpenClaw from its official repository (github.com/openclaw/openclaw). The README also links the official Chinese tutorial at docs.openclaw.ai/zh-CN.
Pick a use case or connect a channel
Choose a use case from the catalog, or connect Telegram / Discord / Feishu / DingTalk / WeCom.
Copy the prompt into OpenClaw
Copy the prompt from the use case and paste it to your OpenClaw. The README recommends using the English prompts in code blocks as-is, since each has a Chinese explanation above it.
Optional: hand the repo to another agent
If you use another AI agent, point it at AGENTS.md first. When forwarding a single use case to an agent outside the repo, send AGENTS.md along with it, because the protocol is not in the use-case files.
Pros & cons
Pros
- Pro:50 use cases grouped by scenario, with difficulty ratings from ⭐ (copy-paste) to ⭐⭐⭐ (needs technical background)
- Pro:China-ecosystem coverage: Feishu, DingTalk, WeCom, Xiaohongshu, WeChat Official Accounts, A-share monitoring, plus China adaptations of international cases
- Pro:Uniform use-case format (pain point, features, required skills, setup with copyable prompts, tips) and a beginner concept table
- Pro:Agent-friendly structure: AGENTS.md, INDEX.md with risk tags, and CONCEPT-MAPPING.md for cross-agent reuse
Cons
- Con:Referenced skills and third-party dependencies are not reviewed by the maintainers; users must check source code and permissions themselves
- Con:The README is a catalog; the actual setup steps and prompts live in the individual use-case files
- Con:Agent-assisted reproduction varies by use-case complexity, and complex ones still need a full read-through